Methodological updates in the Conventional Gait Model 2 preserve kinematic reliability in asymptomatic and cerebral palsy individuals

Three-dimensional gait analysis is widely used to support clinical decision-making in neuromuscular disorders, with the Conventional Gait Model (CGM) being the most commonly applied biomechanical model in clinical practice. Recent developments of the CGM, grouped under the open-source CGM2 framework, introduced methodological updates intended to improve robustness while preserving backward compatibility. However, the reliability of these successive CGM2 iterations has not been comprehensively evaluated, particularly in pathological gait populations. This study investigated within- and between-assessor reliability of lower-limb kinematics across three CGM2 versions (2.1, 2.2, and 2.3) in asymptomatic participants and individuals with cerebral palsy. Reliability was quantified using standard error of measurement and minimal detectable change across the gait cycle. Overall measurement error remained low and consistent across models and participant groups, with standard errors close to 2{degrees} and minimal detectable changes around 6{degrees}. Introducing kinematic fitting had minimal influence on reliability, while adding tracking markers on the thigh and shank produced a modest reduction in hip transverse rotation error. These findings indicate that methodological refinements implemented in CGM2 preserve the reliability of the original CGM while providing incremental improvements for clinically relevant parameters, supporting its use in both asymptomatic and pathological gait analysis and longitudinal clinical assessments